Mecanum drive quickly drains battery

Our drivetrain uses 6" mecanum wheels (AM-3479), VexPro 2 CIM Ball Shifters (no 3rd stage), and 8 Mini-CIM motors. When performing ‘Driver Training’ we are draining our batteries to less than 10 volts in just a couple minutes. We have used both last year’s batteries as well as new batteries purchased for this year’s competition. Robot is loaded to near competition weight (~140#) during these tests. New drivers tend to do a lot of strafing, also we have the SRX brake function enabled.
We have seen teams at the National competitions with similar drive trains that work quite well. Any suggestions on cause/resolution of quick power loss would be appreciated.

Well there’s your issue. I don’t know how many other mechanum teams power their wheels, but in the past 1023 has used 1 CIM per wheel. You should look into the threads about CIM vs MiniCim: the data should be applicable if you’re driving forward.

-Check your main power connections on the robot and batteries
-Feel each motor and determine if they’re heating up evenly or not
-Ensure that the gearboxes were assembled properly and greased
-Push the robot by hand with things powered off (does it roll nicely or is something acting weird?)
-Ensure that your chargers are working properly
-Check every battery with a Beak or similar load tester
-Check that your mecanum wheels are all nice and square to each other

We used a 4Cim+4MiniCim mecanum drive in 2015 and would get 10-20 minutes of acceptable practice drive time on a battery. You are right to be suspicious of something being awry.

You could check the drivers station logs for the current draw of each PDP connection and see if there is a particular motor or motor pair that is drawing an unusual amount of power (relative to the other motors).
With the log viewer you can view a graph of each individual connection.

What are your gear ratios? Are you driving in high or low gear when maneuvering (low recommended)? Are you sure that all of the gearboxes are shifting at the same time; mismatches could cause large current draw from the ones geared faster.

Thanks, these were good suggestions. Most of these have already been verified and check out OK. Will verify remaining items this weekend.
We did not have the problem when we ran the chassis (60# w/o battery), only when we loaded to about 140# total weight.

How did you determine the desired gear ratio? 2 CIM ball shifters with no 3rd stage reduction and 6 inch wheels is beyond aggressive.

Try changing switching to “coast.” Without getting too far into the weeds of how an H-Bridge speed controller works, “brake” will dump the inductive energy stored in the motor windings straight into heating the motor on every single PWM cycle at 10-20 kHz. That’s a big deal if you’re not running your drive motors close to full duty cycle (which happens relatively often for drivetrain motors).

In contrast, “coast” dumps the inductive energy back into the battery every time the PWM switches the H-bridge off, meaning your motors will run cooler and your battery will last longer.

8 MiniCIMs is comparable in power to 8*(2/3) = 5.3 CIMs. That in and of itself isn’t the problem.

I also would like to hear what gear ratios / final speeds you’re using. Checking all your electrical connections is probably a good idea. A DS log graph showing the PDP current from each drive wheel and battery voltage could also be helpful in diagnosing this problem, if you can post that.

I didn’t see this suggestion yet. Have you reviewed the orientation of the Mecanum wheels? There’s a diagram on the AndyMark site that demonstrates the correct orientation of the wheels so that the force vectors are correct. The diagram included is from the top of the robot looking down.

If this orientation is not followed correctly, large amounts of energy will be wasted during certain motions. We ran into a symptom similar to what you’ve described a few years ago. When the problem existed, the ~angle of each roller axis~ (looking from the top down) looked like this:

(“O” configuration):

/
/

What works (and what is documented in the link above) is this configuraiton:

(“X” configuration):

/
/\

  • scott

A 2-stage ball shifter has ratios of 3.667:1/8.333:1. I know from experience that an ~8:1 mecanum setup with 1 CIM per wheel holds up just fine. 5:1, on the other hand, doesn’t. When we tried that, you could drive forward/back just fine, and turn at speed (more like drifting around the turns in Mario Kart), but you couldn’t strafe at all, everything stalled out.

If you’re trying to do anything other than go straight with a little turning control while doing so at high speed, that could explain the problem - the motors are probably near stall and drawing an impressive amount of current (grab the logs and check it, post a screenshot for us!).

Carefully check each roller on each wheel. Do both of these inspections for each roller:

  1. Does each roller spin freely (like it keeps spinning for a bit after you flick it) ?

  2. Do you feel a lot of friction when trying to rotate the roller when applying some axial force on the roller (i.e. slide the roller along its axle until it stops, then push it against the stop while turning it. do this in both directions along the axle) ?

If the answer to (1) is “no”, or the answer to (2) is “yes”, then you’ll have a lot of friction when strafing.

Then put the robot on the carpet and drive it with a straight forward command and record the current draw for each wheel while doing so. Repeat for a pure strafe command. Post your results on this thread.

This is the reason our team moved away from mecanum, we had major issues with battery last year and don’t want the same to happen again.

Did you inspect your mec wheels after assembing, as suggested in this post?

This. Let us know your ratios. You’re likely geared too high for your wheel size.

I have it on high authority that there is no “h” in mecanum.

That strongly supports the hypothesis of incorrect gear ratio and excessive roller friction.

Using the JVN calculator for the 2-CIM Ball Shifters with no 3rd stage gearing, on a 6 inch wheel the lowest speed you can get in low gear is 13.62 ft/sec with a high gear of 30.96 ft/sec. This may be where your problem lies. If you are running in high gear at that fast you are going to have problems. If you are running in low gear then you should be alright for a while, but is still not desirable. I would HIGHLY recommend you add a “3rd Stage Option 2” from Vex which is an additional 30:54 reduction. This will give you speeds of 17.2 ft/sec in high and 7.57 in low.

No, the braking mode is only engaged when the commanded speed is zero (or more precisely, in the deadband around zero). It is certainly possible that if the robot is performing compound maneuvers (e.g. translating at an odd multiple of 45 degrees) that it could be engaged when you don’t really want it, but it’s unlikely to be causing the effects described. Still, couldn’t hurt to check.

Uh…

Your gearing is totally unreasonable; your high gear is 3.66:1. With 6’’ wheels, you’re going to be hard-pressed to move at all without constant brownouts.